基于统计特性的隐写密钥恢复方法  

Steganographic Key Recovery Method Based on Statistical Characteristics

作　　者：郭微光 李佳临 徐明迪[2] GUO Weiguang;LI Jialin;XU Mingdi(No.91001 Troops of PLA,Beijing 100841;Wuhan Digital and Engineering Institute,Wuhan 430205)

机构地区：[1]91001部队,北京100841 [2]武汉数字工程研究所,武汉430205

出　　处：《计算机与数字工程》2023年第5期1113-1119,共7页Computer & Digital Engineering

摘　　要：隐写分析技术是检验信息隐藏技术安全性的关键技术,作为隐写分析的最终目标,隐秘信息的提取在取证隐蔽通信方面有重要作用。现有明文嵌入条件下的密钥恢复方法会将部分伪密钥误判为真,为避免这种情况,论文提出了一种基于明文统计特性的隐写密钥恢复方法。该方法在估计嵌入率的基础上确定编码参数的范围,构造密钥空间;然后利用校验矩阵和载密序列通过STC(Syndrome-Trellis Codes)译码方程求出译码序列;最后,基于明文信息的统计特性对隐写密钥进行识别。实验结果表明,针对自适应隐写算法,该方法能够盲识别出正确的隐写密钥。与现有的盲识别算法相比,该方法提高了隐写密钥识别的准确率。Steganalysis is the key to test the security of information hiding technology.As the ultimate goal of steganalysis,the extraction of covert information plays an important role in forensic covert communication.The existing key recovery methods under the condition of plaintext embedding may misjudge some false keys as true.To avoid this situation,this paper proposes a steganographic key recovery method based on the statistical characteristics of plaintext.This method determines the range of encoding parameters on the basis of estimated embedding rate to construct key space.Then the decoding sequence is obtained by syndrome-trellis codes(STC)decoding equation using check matrix and stego sequence.Finally,the steganographic key is identified based on the statistical characteristics of the plaintext information.Experimental results show that this method can blindly identify the correct steganographic key for the adaptive steganography algorithm.Compared with the existing blind identification algorithms,this method improves the accuracy of steganographic key identification.

关 键 词：自适应隐写 统计特性 参数识别 密钥恢复 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]